[_7gt4]

Mastodon instances are not public spaces. Users can pick how much they want to engage in political discussions by picking their instance or, if they can, hosting their own. (The second part of the argument is flawed because most people can't just throw up a rails app and maintain it, be it cost or lack of knowledge)

What most people don't get about "bubbles", mostly because it doesn't affect them, is that sometimes, if your mere existence is political, getting some rest from it can be quite hard.

[unstuckdev]

Some people confuse boundaries with bubbles.

When they tell me I'm living in a bubble because I use an instance with rules and instance blocks, what they want is for the street preacher to be able to chase me down the sidewalk yelling fire and brimstone at me. "I wouldn't personally do that," they insist, "but people should be able to!"

They want to be able to harass people online the way they do offline, and on less advanced social networks. It's as ideological as all the opinions they reject as ideology.

[makomk]

One of the ways instance blocks are used, in practice, is to discourage other instances from enforcing boundaries. For example, one of the biggest and oldest Mastodon instances recently blocked an entire up-and-coming instance because they banned one person who has a trail of allegations that they solicited underage kids for sex. A number of other, smaller instances also block instances that ban another alleged serial rapist.

[loceng]

Does the functionality of blocking just a single user from an instance not exist with Mastodon?

This does bring up an important point however, in that the largest instances that will exist may not be the one that curates and manages the instance/ecosystem/users on the ecosystem as well or as reasonably, however creates a space (or bubble) of popular reaction - say being only "80%" compassionate vs. fully understanding. For example, what if a convicted rapist has done their prison time? I feel this would have very different responses based on the nation and culture and how much or how little actual rehabilitation and treatment people receive, if that society trusts the outcome of the system or not. I believe that keeping "unhealthy" people separate from "healthy" people will prevent them from learning - and of course has the potential to allow the unhealthy to learn further bad behaviours or reenforce ones they already have; this doesn't mean we should design for a free-for-all system like Twitter and Facebook are various degrees of. It comes down to compassion - which includes not dismissing or forgetting about people, even if they're not people we'd want as our friends or family.

In the case of a user who simply has a "trail" of accusations for unacceptable behaviour, do we foster "guilty until proven innocent" as an acceptable behaviour - blocking that user from society's online interactions - or should we foster more of a "innocent until proven guilty" - while perhaps keeping an eye on people who are accused of certain behaviour? If the a little girl accuses her brother of stealing the cookie from the cookie jar, however the boy claims it didn't happen - and there's no proof, it's not fair to me if to punish either child until there's proof: either the sister took the cookie and is lying, or the brother took the cookie and is lying; you could ban them from the kitchen with the cookie jar, though if they're ever left alone and there's no evidence like video to check back on, then either child could still take another cookie.

I then wonder though, if there become more trusted and larger instances that do put the effort (and cost) into curation and managing the community (the family of society), then smaller instances will and can tap into that network - however then not contributing or covering any of the cost. How does this imbalance balance out, or does it need to, even? Does it become a question as to whether people who want accountability are willing to pay a small amount to pay the costs of such moderation vs. people simply hoping someone else deals with it - a bystander effect?

I wonder if the Mastodon community or founder have any thoughts or beliefs to this? Likewise, what happens if you have a whole bunch of instances where people are in their bubble of communities, perhaps with less critical thinking time spent and more indoctrination, where then propaganda can easily be distributed without any potential to filter it out and protect the whole as easily; of course solid in-person, real-life community and connectedness is the only way to counter this, including developing deep trust with individuals who you trust are thoughtful and such.

I'm still not sold on that decentralized is the safest option, however it perhaps can help counter against bad actors who seek control of systems; in fact it could act as a canary if the powers that be attempt to make decentralized systems illegal - as part of an effort to control (and perhaps censor). However, so long as the community is educated as a whole, and is hyperaware of these characteristics and readies to counter it in full, to hold the line for peace and justice, then we're on the right path. In contrast to decentralization, if we look at Apple's ecosystem, trust and governance and good design, forethought, plays an important role in success - though everyone in that ecosystem is "paying their dues" via profit Apple receives. And contrasting that further to Facebook, whereby they profit off of the manipulative aspect of ads and with practically no vetting or oversight.

[_7gt4]

this is a huge wall of text and I haven't necessarily read it all because I'm tired (I'll bookmark this I swear) but the case mentioned is a person with repeated persistent confirmed-by-lots-of-people manipulative bad behavior with at least two independent rape accusations on top of that.

If you're building a purely general purpose instance or a free speech themed one it's fine to not block this sort of instance.

It's okay to have this "net split". It's a system of curation. It encourages smaller instances.

Mastodon is also heavily used by minority / LGBT communities that deal with abuse and harassment a lot, often for merely existing. And often a "free speech" instance or worse a "no moderation" instance will attract that sort of person.

[makomk]

While I guess there are arguments for not banning users just because they have a whole bunch of nasty accusations against them, or for requiring high standards of evidence to do so, that wasn't the actual justification from what I can tell. (Indeed, I think arguing that rape accusations shouldn't automatically be believed is probably a bannable offence on all the instances involved.) It's more like they don't think community insiders should be held to the same standards as other people, and reckon any outsiders who do so are so obviously in the wrong as to justify banning the entire instance.

[sdrothrock]

I'm not only thinking about political bubbles here, but linguistic or cultural ones. For example, when I joined a server here, it was federated only with a few other small Japanese servers, resulting in a fairly miserable experience. Another set had multiple languages but was just devoted to art. It's incredibly difficult (or was a year ago) to find servers that are general purpose and federated with both Japanese and English servers.

It's actually kind of interesting that me mentioning bubbles resulted in an immediately political assumption; I guess it shows how out of touch I am with mainstream Twitter.

[detaro]

I think you use "federated with" in a different sense than most people do in these discussions.

A newly created Mastodon instance by default federates with everyone in that allows messages and following from and to every other instance.

But it's federated timeline has no posts from other instances, because that view only shows posts that the instance actually sees. It only shows what people on the same instance actually follow, because otherwise there is no reason for the posts to be sent to it.

[sdrothrock]

That would explain a lot of things and also clears up a lot of things for me, too! Thank you. I guess a better phrase for what I'm trying to describe is a lack of discoverability, then.

[_7gt4]

I only know one instance (awoo.space) that does whitelisting instead of blacklisting. Mastodon can't even do it out of the box (although pleroma can).

My instance bans the sort of instance described in the sibling comment to yours, like qoto (who advertised their Mastodon as "free speech twitter"... on twitter ads)